A vacation-rental ad account sat dark for nine days. Now it's checked every day.
An Ontario vacation-rental company's Google Ads recorded every booking request at $0, spent on search terms that never produced one, and stopped running for nine days before anyone noticed. Here is what I changed, what it did, and what it does not prove.
The short version. An Ontario vacation-rental company runs one Google Search campaign. When I started running it, every booking request was recorded at $0, some search terms kept spending without ever producing one, and the ads had been dark for nine days. I relit the campaign the day I found it, and it has served every day since, 125 of 125 through Oct 12.
A check now runs on the account every scheduled day, and every tracked booking request I have checked against the owner's own records matched. What it does not prove: more bookings, or revenue. Booking requests are about level with last year, and the page says so.
The ads were spending. Nobody could see what for.
The campaign was running and requests were coming in, but three things were hidden. A handful of search terms were each taking $25 or more without ever producing a booking request. Every request that did come in was recorded at $0, so a quiet midweek inquiry and a four-figure peak-week request looked identical. And nobody was reading the account day to day.
That last one had already cost something. A leftover campaign end date had expired, and spend sat at $0 for nine days heading into summer while the status still read Enabled. I found it by reading the account, and brought it back the day I found it.
Relit it, cut the waste, fixed the counting, then kept watching.
The ads. I blocked search terms that kept spending without producing requests, rewrote the ads so regional searches see regional wording with the lakes and areas people actually search for, and added photos to the ads. None of it needed a bigger budget.
The tracking. Booking requests from the ads now count in the ad account once the booking system records them, carrying the total the guest asked for instead of a zero. That is the requested value, not collected revenue, and the campaign still bids on the number of requests and calls, not on that value.
The watching. After the nine dark days, I built an automated check that runs every scheduled day, looking for a dark campaign, a spend stall, or broken tracking. That check is now a service on its own: website monitoring.
The reporting. The owner has a live dashboard and a plain-language weekly brief.
What changed, with the limit beside each number.
Figures as of October 12, 2026, shown anonymized; the check's run count is dated in its tile. Each one says what it is measured against, because a number without its window is how case studies mislead.
Days the ads have served since I relit the campaign.
Jun 10 to Oct 12. Before that, the ads had sat dark for nine days, which I found by reading the account.
An automated check is scheduled on the account every day.
Built after the nine dark days. I can show you the run log.
Booking requests from the ads now count in the ad account once the booking system records them, carrying the total the guest asked for.
Requested, not collected. The owner confirms and takes payment offline, so it is a reporting signal, not revenue.
Tracked website booking requests I checked against the owner's own booking system, one at a time. All of them matched, amount included.
Aug 1 to Oct 9, checked at each month close. It shows the tracking is right. It does not show more bookings.
What these results do and do not show.
Requests are not revenue. The booking system records booking requests, not payments, so no ad click can be joined to a paid stay. I do not claim a revenue figure, a return on ad spend, or that the ads filled the calendar.
Better measured is not more bookings. Booking requests from June 9 to October 11 were about level with the same weeks last year, and requests from the ads fell through late summer. Some of that is the season, since late summer is this account's quietest stretch for bookings. Not all of it is explained yet, and I am not going to pretend otherwise.
No season-against-season numbers. This page used to compare a spring stretch with a summer one. Clicks are cheaper and searches busier in summer, so that comparison flattered the result. I took it off on October 15.
What happens next. The first fair test is the January to March 2027 booking season, against the same months this year. I will update this page with whatever it shows.
Questions about this engagement.
Did the ads actually get better, or just better measured?
Better run and better measured, and I can show both. The campaign has served every day since I relit it, a check runs on it every day, and booking requests now carry what the guest asked for. More bookings is not proven yet. Booking requests from June 9 to October 11 were about level with the same weeks last year, and requests from the ads fell through late summer. Comparing spring with summer would flatter the numbers, so I do not. The first fair test is the January to March 2027 booking season.
Did this make the business more money?
I cannot prove that, and I will not claim it. The booking system records requests, not payments, so there is no way to join an ad click to a paid stay. What I can show is an account that serves every day, is checked every day, and records what each request was worth.
Is the requested booking value real revenue?
No. It is the total the guest asked for in a booking request, not collected revenue, since these bookings are confirmed offline. It is a reporting signal for the owner. The campaign bids on the number of booking requests and calls from the ads, not on that value; any move to value-based bidding is a separate, evidence-gated decision.
Want to know what your ad account is actually doing?
Automation Solutions Co. · Uxbridge, Ontario · replies within one business day